[erlang-questions] node.js vs erlang

[zxq9]

On Thursday 19 June 2014 09:05:43 Karl Nilsson wrote:
> I followed the cowboy getting started tutorial and found it very reassuring
> to be taken all the way to generating a release. I am pretty sure I could
> build on that sample and actually create something that could be used. I
> think there is a lot of value in that. Personally I find tweet sized
> getting started examples very unsatisfactory and prefer something that
> gives and indication of workflow as well as the language. Hipsters may feel
> differently.

Django users certainly don't. The Django documentation and the Python 
documentation -- two non-trivial documentation projects if there ever were -- 
are primary examples of how to speak directly to an open source community in 
its own language while making things approachable without dumbing anything 
down.

The Postgres docs are another stellar example of documentation, notable 
particularly because they are *not* written for the average grade-school 
student and have only become popular very, very slowly as (a few elements of) 
the web dev crowd have matured and realized that mysql doesn't fit every 
purpose.

The Erlang docs already fall into the Postgres category. The challenge is to 
explain the *runtime* and its capabilities in a way that compares to the 
Django docs -- in particular, how to think of a problem in terms of 
independent, concurrent processes (instead of getting bogged down in the 
management of concurrency the way teaching the same ideas in C/Java/Python/etc 
would) and how to leverage the Erlang system.

The Erlang language itself does not need much explanation. Teaching basic 
functional programming paradigms is not the best way for the Erlang community 
to use its time -- RWH, LYAH, LYSE, and SICP already do an excellent job of 
that. After going through any one of those Erlang is readable.

This whole discussion is still not separating Erlang the language from Erlang 
the platform, and it must if we want to understand how popularity works -- and 
let's be clear, this whole discussion is an envy-driven brainstorm about how 
ubiquitous low-tech could possibly trump something of clear technical 
superiority.